3.4.1. Principles of the CA–Markov Model

The Markov chain model is a Markov model with a discrete time and state. It is often used to analyze the law of land-use changes. Cellular Automaton (CA) can be used to simulate the process of natural changes with the influence of the surrounding space. In this study, CA was utilized to simulate changes in any pixels in the land-use image affected by the state of themselves and their neighboring pixels.

The traditional Markov model has difficulty predicting spatial changes in land-use patterns. The CA model focuses on partial interactions between cells, which has obvious limitations [23]. The CA–Markov model has the advantages of both the Markov model and the Cellular Automata, as spatial neighboring elements and rules of spatial distribution conversion are added to the Markov chain model. It can be used to predict spatial changes of the land-use pattern with full consideration of spatial parameters.
